At present, there are many duplicate checking softwares, and the detection rules of different softwares are different. Combined with various detection software, the general law of repetition rate detection is simply sorted out, which can provide some reference for those who have the demand for paper creation.
What is plagiarism? Take the most widely used HowNet as an example. Its detection method adopts the most advanced fuzzy algorithm at present. It has a premise and conditions. Usually both are regarded as plagiarism or suspected plagiarism.
1, one premise: the segmentation is given a threshold of 5%.
2. One condition: 13 consecutive characters are the same.
What do you mean? Let's illustrate, for example, that if a paragraph quotes 13 words from other original texts, if this paragraph * * * has 100 words, because the quoted words account for 13%(> 5%), it will be detected as plagiarism. If this paragraph has 400 words, then the quotation accounts for 3.25%.
Of course, different systems have different algorithms and rules, and there are different views on which system is stricter, but I just want to remind you of the following two points:
First, the paper should be original, and the research methods can be used for reference, but the results of predecessors cannot be copied;
Second, when testing papers, especially graduates of master's degree and junior college, we must know what kind of testing system our school uses, and choose the system and version consistent with the school for testing. Spending more money and doing less is really not worth the loss, which will affect graduation and degree.
References:
PaperPP paper repeated inspection